Enhancing interventions for prevention of mother-to-child- transmission of hepatitis B virus

Summary Prevention of mother-to-child transmission of hepatitis B virus (HBV) infection is a cornerstone of efforts to support progress towards elimination of viral hepatitis. Current guidelines recommend maternal screening, antiviral therapy during the third trimester of high-risk pregnancies, universal and timely HBV birth dose vaccination, and post-exposure prophylaxis with hepatitis B immunoglobulin for selected neonates. However, serological and molecular diagnostic testing, treatment and HBV vaccination are not consistently deployed, particularly in many high endemicity settings, and models predict that global targets for reduction in paediatric incidence will not be met by 2030. In this article, we briefly summarise the evidence for current practice and use this as a basis to discuss areas in which prevention of mother-to-child transmission can potentially be enhanced. By reducing health inequities, enhancing pragmatic use of resources, filling data gaps, developing advocacy and education, and seeking consistent investment from multilateral agencies, significant advances can be made to further reduce vertical transmission events, with wide health, societal and economic benefits.


Background
Prevention of mother-to-child transmission (PMTCT) of hepatitis B virus (HBV) infection is a cornerstone of strategies for elimination of viral hepatitis. Specific international public health targets for HBV include 95% reduction in new infections in children, 1 and lowering prevalence in 0-4 year olds from 0.9% in 2020 to 0.1% by 2030. 2 PMTCT is crucial to tackle HBV infection and disease, as perinatal infection is associated with a >90% chance of chronic infection, 3,4 with associated long-term risks of liver disease, including hepatocellular carcinoma (HCC) in young and middleaged adults. Between one-third to one-half of all chronic HBV infections have traditionally been attributed to vertical transmission, but this proportion is increasing over time as a result of vaccination programmes that improve immunity in older age groups. 5 Neonatal acquisition is now the major cause of chronic hepatitis B worldwide.
Thus, there is now an urgent need to review and revise practical, evidence-based approaches for HBV PMTCT. We set out to provide a summary of current evidence underpinning PMTCT recommendations, in order to highlight challenges, identify evidence gaps and promote potential advances in strategy. Elimination challenges differ between geographic regions, so lessons must be learned from existing experience and interventions tailored to find the best solutions for individual local contexts, while universally promoting efficacious, equitable and evidence-based interventions. Our approach aims to promote expert debate in areas that remain unresolved, inform further data collection and research, and provide advocacy for interventions, policy and funding to support best practice.

Search strategy
To support our narrative review and inform opinions, we initially identified references through clinical guidelines and searches of PubMed, prioritising publications since 2015, but drawing on older primary literature where required for an evidence base. We used the search terms "Hepatitis B" and "HBV", together with "mother to child transmission", "mother to infant transmission", "perinatal transmission", "vertical transmission", "elimination", "pregnancy", "birth", "in utero", "vaccination". Material was also identified from the reference lists of included articles, conference abstracts and through searches of the authors' own files. Only papers published in English were reviewed. The final reference list was generated on the basis of originality and relevance to the broad scope of this review.

Risk reduction through PMTCT interventions
PMTCT recommendations are set out by the World Health Organization (WHO), 6 and in regional HBV guidelines for Europe, 7 the US, 8 and Asia Pacific regions (Table 1). 9 Interventions combine screening of pregnant women, antiviral prophylaxis with nucleos(t)ide analogues (NAs) for highrisk mothers from 28 weeks of gestation onwards, infant birth dose (BD) vaccination, and one or more doses of hepatitis B immunoglobulin (HBIG) administered to high-risk neonates, as a component of the WHO 'triple elimination initiative' to reduce the morbidity from HBV, HIV and syphilis. [10][11][12][13] All infants should then subsequently receive a further three doses of HBV vaccine, combined into a multivalent formulation, as part of the routine WHO expanded programme for immunisation.
Rigorous deployment of these interventions makes a substantial impact on vertical transmission rates ( Table 2). Success has been exemplified in China, whereafter an initial partnership with Gavi (the Vaccine Alliance) and subsequent government fundinga combination of improved antenatal HBsAg screening, provision of free BD vaccination, upscaling of the full vaccine schedule and deployment of village lay healthcare workers has reduced HBsAg seroprevalence in children <5 years from 9.7% in 1992 to <1% in 2013. 14,15 However, the implementation of PMTCT measures lag far behind recommendations in many settings. There is poor awareness, lack of access to affordable tests for diagnosis and risk assessment, and limited availability of antiviral drugs, vaccines and HBIG, while the COVID-19 pandemic has disrupted existing immunisation programmes, widened health inequities, amplified vaccine hesitancy, and diverted attention and resource from HBV. [16][17][18][19] Maternal testing and prophylactic antivirals: current practice Screening and risk stratification Worldwide, most people living with HBV infection are not aware of their status; 20 this also applies to pregnant women. Screening for HBV is recommended in all pregnancies, 6 using a point of care test (POCT) 21 or a laboratory ELISA to detect hepatitis B surface antigen (HBsAg) from venous blood or dried blood spots (DBS), but the advent of universal HBV vaccination has diminished the incentive for HBsAg testing of pregnant women in many regions of the world. HBsAg POCT has the advantage that follow-up samples can immediately be collected to stratify risk and inform management. This typically includes: (i) HBV DNA viral load: A higher risk of vertical transmission is associated with maternal HBV DNA >200,000 (5.3 log 10 ) IU/ml 22,23 despite HBV vaccine series plus HBIG, and thus antenatal maternal NA prophylaxis is also recommended. 6,24 (ii) HBV e-antigen (HBeAg): In situations where HBV DNA measurement cannot be undertaken, positive HBeAg is a proxy for viral load >200,000 IU/ml, with a reported sensitivity of 88% (95% CI 83.9-91.5) and specificity of 93% (95% CI 90.0-94.5). 24 A decline in HBV DNA to <200,000 IU/ml by the time of delivery occurs in the majority of women who start NA prophylaxis at 28 weeks gestation. 25 Data from randomised controlled trials (RCT) of tenofovir disoproxil fumarate (TDF), based on >1,300 pregnancies, reports MTCT rates of 0-6% in those receiving antiviral prophylaxis, compared to 2-30% in controls, 26 while formal meta-analysis of RCT data generated pooled odds ratios of between 0.10 and 0.16 for different NA agents (Table 2). 27

Safety of antiviral therapy in pregnancy
The US Food and Drug Administration (FDA) 28 reports no increased risk of congenital abnormalities associated with TDF, telbivudine (LdT), or emtricitabine (FTC) in pregnant women with HBV infection. A systematic review of the safety of antiviral HBV drugs in pregnancy collected data on neonatal outcomes (death, prematurity and congenital abnormalities, and also bone mineral density for TDF) and maternal outcomes (miscarriage/ stillbirth, post-partum haemorrhage, antiviral resistance, and hepatitis flare), and found no evidence of increased risk for any of these adverse events. 29 Tenofovir alafenamide (TAF) was approved for the treatment of HBV in adults and adolescents in 2016. TAF delivers the same active tenofovir metabolite as TDF, but at lower doses, thus reducing the incidence of side effects. 30 Although TAF is thought to have a reduced renal and metabolic bone side-effect profile compared with TDF, it is potentially associated with weight gain, altered lipid profiles, and increased cardiovascular risk. 31,32 However, in HBV PMTCT studies using TAF from gestational week 24 onwards, there were no significant side effects or toxicity, and no vertical transmission events. 33,34 During breastfeeding, negligible concentrations of drug have been measured in breast milk, supporting its safety in ongoing treatment post-partum. 35 The ALLIANCE trial reports an increased rate of favourable HBV endpoints (HBsAg loss, HBeAg seroconversion, alanine aminotransferase [ALT] normalisation) in previously treatment-naive HBV/HIV-coinfected individuals receiving a TAF regimen compared to a TDF regimen. 36 However TAF has not yet been approved for HBV PMTCT and more data are needed. Furthermore, the high price of TAF means it is not yet a cost-effective choice even when healthcare is wellresourced. 37 Experience accumulated through HIV ART also underpins important evidence for NA administration early in pregnancy, and provides information about the efficacy and safety of combination therapies (not well studied in HBV cohorts to date). A systematic review that collated outcomes on >19,000 pregnancies reported elevated risks of preterm delivery and lowbirthweight infants when ART was started pre-conception vs. during pregnancy; however, this result may be confounded by other factors, and there was no evidence of other adverse outcomes, leading the authors to conclude that the overall benefits of ART outweigh the risks. 38 A registry-based study in South Africa found no increased risk of congenital malformations in pregnancies exposed to first-line ART (including TDF/FTC) started prior to pregnancy vs. later, 39 and in Botswana, this regimen was associated with better infant outcomes. 40 Follow-up of >3,500 children in the US who were exposed to ART in utero did not identify any significant increase in adverse outcomes, up to the age of 13 years, 41 corroborated by a more recent study of the impact of first-trimester TDF. 42 Despite a possible association with a reduction in mean whole-body bone mineral content in the newborn, 41 a double-blind RCT reported no evidence of a TDF effect on bone density in either mother or infant at 1 year of age, 43 and other studies have also provided reassuring data. 44,45 Maternal interventions: challenges and data gaps Access to screening and stratification Implementation of maternal risk assessment and antiviral therapy is currently impeded by inconsistency in routine antenatal HBsAg screening, and poor access to HBV DNA and/or HBeAg testing. Even when appropriate consumables and clinical/laboratory facilities are available, the costs of additional tests are often prohibitively high, either to the patient (where cost is outof-pocket), to the state healthcare system (when there is government provision), and to private insurers (who may therefore not agree to cover costs).
HBeAg is typically cheaper than HBV DNA and more likely to be accessible, but it is not always a reliable proxy for HBV DNA quantification, particularly for certain viral genotypes. [46][47][48] Thus, among those testing HBeAg-negative, 12-16% of women typically have HBV DNA > − 5 log 10 IU/ml and 2% have HBV DNA > − 7 log 10 IU/ml, 24,27,29 but the relationship between HBV DNA levels and HBeAg status varies between settings, at least partly dependent on the prevalent viral genotype (e.g. for genotype E, HBV DNA may be high in the absence of detectable HBeAg, while in other populations HBeAg positivity is associated with lower HBV DNA levels). 24,49,50 Furthermore, HBeAg is subject to genomic, transcriptional, translational and post-translational changes. Other markers have been considered as potential substitutes for HBV DNA testing, including quantification of serum HBsAg (qHBsAg), or core-related antigen (HBcrAg) which correlates closely with HBeAg. 51 However, these measures correlate imperfectly with HBV DNA (the picture is confounded by the fact that HBsAg is derived from both covalently closed circular DNA and replicating virus, but also from integrated viral genomes). 52 Such tests are not yet widely available and more data are needed to determine their utility in PMTCT risk stratification.
In order to improve reliable and equitable risk stratification and reduce dependence on HBeAg, a rapid, affordable and POCT (or near-POCT) for HBV DNA is urgently needed. Infrastructure established for management of HIV infection has demonstrated that viral load testing can be made widely available. An inexpensive laboratory or POCT for HBV DNA could be added as a reflex follow-up to a positive HBsAg test, or could replace HBsAg screening, conferring the benefits of a single test that can be used for both diagnosis and risk stratification. A high sensitivity test is not required; a POCT for HBV DNA could be developed to provide a readout at the agreed threshold for treatment, with or without quantification over the range that informs use of antiviral prophylaxis and treatment. Long-term, broader access to testing for HBV DNA also allows for identification of cases of occult HBV infection (defined as cases in which HBsAg is negative, despite HBV viraemia). While occult HBV infection is not thought to be a significant contributor to MTCT, as HBV DNA concentrations are typically low, occasionally viraemia is at levels that are clinically significant and thus require treatment or prophylaxis.
In addition to HBV DNA testing, POCT for other biomarkers are also available or under development, with tests for detection and/or quantification or semi-quantification of HBeAg, HBsAg, HBcrAg and ALT. Careful validation of performance, accessibility, economic feasibility, and impact on care delivery is needed. A list of 'prequalification' tests has been produced by the World Hepatitis Alliance. 53 Treatment eligibility based on fibrosis scores In individuals with evidence of significant liver fibrosis or cirrhosis associated with HBV infection, antiviral treatment should be started and continued irrespective of pregnancy (Table 1). However, identifying this group is challenging, as access to elastography is limited in many settings due to the high cost of hardware and software, alongside staff training and ongoing maintenance costs. Furthermore, elastography is not validated in pregnancy. Alternatively, estimation of fibrosis can be derived from laboratory parameters, such as APRI (aspartate aminotransferase-to-platelet ratio index) or FIB-4 (Fibrosis-4), but such scores have variable sensitivity and specificity, and further exploration is needed in diverse populations. 54,55 A school of thought exists to suggest that infectious levels of virus observed in pregnancy may be an indication to continue treatment of the mother in her own right, irrespective of the level of serum aminotransferases or stage of fibrosis.
HBV DNA thresholds for antiviral prophylaxis Although a quantitative threshold has been determined for antenatal prophylaxis, in practice there is a progressive increase in MTCT risk with rising maternal HBV DNA levels, estimated as an odds ratio of 3.49 (95% CI 1.63-7.48) for each log10 increase in concentration, after adjustment for other factors. 22 This observation, at least in part, accounts for the elevated risk of vertical transmission in younger mothers, 56 in whom rates of HBV replication (and HBeAg) are higher. MTCT risk is also higher if HBV DNA is found in cord blood, 57 and when acute maternal HBV LdT, telbivudine; MTCT, mother-to-child transmission; PMTCT, prevention of mother-to-child transmission; TDF, tenofovir disoproxil fumarate; WHO, World Health Organization. HBIG guidelines are all based on post-partum IM administration of HBIG (0.5 ml), to be administered in combination with HepB-BD vaccine (though recommended at a different site). * Many documents refer back to older WHO/CDC guidance from 1984, 121 which pre-dates antiviral drugs and vaccination, to give HBIG preferably within 12 h of birth for HBeAg-positive mothers, with repeat HBIG at 3 months and 6 months to further reduce the probability of chronic infection.   86 3,429 participants (mode of delivery analysis); 2,443 participants (mode of feeding analysis) Caesarian section and nonbreastfeeding In the absence of maternal antiviral prophylaxis, MTCT risk: (continued on next page) Review 6 infection occurs during pregnancy. 58,59 Lowering thresholds for antiviral prophylaxis in pregnancy would reduce the risks of significant viraemia at term. The current HBV DNA criterion for maternal antiviral prophylaxis is >200,000 (5.3 log 10 ) IU/ml, but lower thresholds have been suggested, with 10,000 (4.0 log 10 ) IU/ml as an indicative viral load at which infection may occur despite HBV vaccination if given alone without other PMTCT interventions. 60 A lower threshold of viral load >2,000 (3.3 log 10 ) has been proposed as a relaxed eligibility criterion for wider adult HBV treatment (irrespective of liver enzymes or fibrosis stage), and undetectable HBV DNA levels would mitigate any quantifiable transmission risk, based on the 'undetectable = untransmissable' (U=U) paradigm that is now widely applied to HIV. 61 Particularly in settings where gaps in access to HepB-BD vaccine persistconsistent maternal screening and broadened eligibility criteria for prophylaxis could be of value to minimise the transmission risk. More evidence is needed to inform this agenda.
Offering prophylaxis to all pregnant women who test HBsAg positive has been explored. This approach removes the requirement for DNA or HBeAg testing to stratify risk, and provides opportunities to start treatment promptly, reducing the risk of delays or loss to follow-up, redressing health inequities, and preventing lifelong infection; economic analysis suggests this can be a cost-effective approach, 62 though risk-benefit assessments in different populations are required.
Starting and stopping antiviral prophylaxis At present, NA prophylaxis is only recommended during the third trimester of pregnancy. However, the trajectory of decline in HBV viraemia on treatment is such that HBV DNA may not fall below 200,000 IU/ml by the time of delivery in a proportion of the population with high levels of viraemia (>30% in one study, although also impacted by incomplete adherence). 25 Starting antiviral therapy earlier in pregnancy, potentially as soon as maternal HBsAg-positive status is confirmed, may therefore be advantageous in providing a longer period for virologic suppression (recognising that this can take weeks to months), 63 providing better protection for babies born pre-term, and also possibly reducing the small risk of in utero HBV transmission events. 9,64 Initiating TDF at gestational weeks 14-16 can suppress HBV DNA from >200,000 IU/ml to <20,000 IU/ml in 96% of cases, 65 with no MTCT events and no increase in adverse events in mothers or infants. 66 Meta-analyses suggest an advantage to starting lamivudine (3TC) at < − 28 weeks gestation compared to later in pregnancy, 29 (although this agent is now less widely used for HBV due to resistance concerns). Other NA agents can be initiated in trimester one or two, without any adverse impact on safety (more details in Table 2). 67,68 Clinical trials are in progress. 69 Continuing maternal NAs post-partum provides a benefit for mothers in whom treatment is indicated by current guidelines, outside the context of PMTCT. In those who only required NA drugs for PMTCT, guidelines vary, with some recommending continuation for up to 12 weeks post-partum, 7,8,70 while others suggest prophylaxis can be stopped at birth. 6,9 However, in settings where HepB-BD delivery is not universal, stopping maternal medication before the infant receives at least one dose of HBV vaccination as part of multivalent infant vaccination poses a potential risk of horizontal transmission from mother to infant, and thus there is a rationale for continuing prophylaxis at least until this time point. Recommendations for antiviral treatment of women before, during and after one or more pregnancies need to be informed by wider programmatic decisions about HBV treatment in all adults, with consideration of the potential individual-and population-level benefits of lowering the DNA threshold that determines eligibility. [71][72][73] In addition to reducing the risk of long-term liver disease in the individual, future strategies that endorse ongoing maternal antiviral therapy would also reduce MTCT risk for subsequent pregnancies, tackle the risk of horizontal transmission events within households, and abrogate the risk of severe liver flares associated with NA cessation. The optimal strategy will require pre-treatment HBV DNA quantitation, and the infrastructure to provide ongoing clinical monitoring and access to treatment for the mother; achieving this in practice is currently challenging in many settings but must be scaled up.
Potential risks of NA prophylaxis NA agents are generally safe and can be used in pregnancy, although more data would be advantageous to review outcomes of earlier and wider prophylaxis, and to provide better representation of diverse populations. If expanding NA prophylaxis, benefits must be weighed against the administration of medication to mothers with relatively low HBV DNA concentrations (where transmission risk should be mitigated by HepB-BD vaccination plus routine follow-up immunisation from 6 weeks of age). In settings in which neonatal and infant immunisation rates are poor, there is currently an argument for offering antiviral therapy to more pregnant women, while it also remains paramount to continue advocating for improved vaccine access.
There is a risk of hepatitis flares when treatment is stopped post-partum, with modest ALT elevations (e.g. with ALT rise to >2x upper limit of normal) in more than one-third of women, 74,75 of whom some may seroconvert to anti-HBe or even lose HBsAg, and others may become eligible for long-term therapy. Interestingly, these flares may be less common in women who have received peripartum prophylaxis than in their untreated counterparts; 76 reduced flares may be due to continuation of antivirals during the post-partum period. Most episodes of post-partum transaminitis are asymptomatic and resolve spontaneously; 77,78 however, rare serious hepatitis resulting in morbidity and liver decompensation can occur. 79 The risk is somewhat unpredictable, but is higher in younger and HBeAg-positive women. 78,80 Access to a reliable, affordable, and consistent source of medication is therefore required.
Post-partum, it is impossible to gauge which mothers would benefit most from ongoing NA therapy unless HBV DNA quantification is undertaken at baseline. However, if overall guidelines for adult treatment are simplified, with wider treatment eligibility and access, this may help guide stratification and clinical management decisions during and after pregnancy. The cost of generic TDF (frequently <50 USD/per year in many countries) or, in the future generic TAF, should not be a barrier to treatment. [71][72][73] Single agent vs. combination prophylaxis Although TDF is low cost and is included on the WHO list of essential medicines, there is poor access to tenofovir monotherapy in many settings where HBV is prevalent. 81 Advocacy is urgently required to improve drug supply chains, with regional procurement tackled as a priority. Some clinical programmes have pragmatically capitalised on the availability of either TDF or TAF in combination antiviral preparations that are widely available for HIV treatment and pre-exposure prophylaxis, for example TDF/3TC. Furthermore, dual therapy could reduce HBV DNA more quickly and robustly than monotherapy, and emerging data show the potential benefits of combination therapy in chronic HBV infection, 82 though there is no evidence base in pregnancy to date. Studies comparing single agent vs. dual therapy are warranted.

Maternal HBV vaccination
In high endemicity settings, and in pregnant women at active risk of HBV exposure, it has been suggested that those testing negative could be offered HBV vaccination, which is safe in pregnancy and protects against acute infection (itself associated with a high risk of MTCT). 83,84 However, such a programme consumes additional resources focused on the HBV-negative population, who should ideally be screened for anti-HBs and anti-HBc such that vaccination can be targeted at those who lack pre-existing immunity. In populations with a high prevalence of active HBV infection, the prevalence of prior exposure (anti-HBc) in the HBsAg-negative population is high, and routine catch-up vaccination for adults is not regarded as necessary or an appropriate use of resources. 85 Approaches to maternal vaccination should therefore be reviewed based on risk assessment for individual women, alongside local epidemiology to inform population-wide strategies.

Delivery route and breastfeeding
There may be a reduction in MTCT risk associated with Caesarian section vs. vaginal delivery, and in non-breastfeeding mothers 86,87 (results of meta-analyses presented in Table 2). However, these interventions are not suggested for HBV PMTCT, as transmission risks can be adequately mitigated with other interventions, and there are other significant risks for mothers and babies associated with surgical delivery and with nonbreastfeeding.

HepB-BD vaccination: current evidence and interventions
HepB-BD has been incorporated into WHO guidance for all settings since 2009. 88,89 At a cost of 0.25 USD per dose (UNICEF data for 2022), 90 monovalent HepB-BD vaccination is deemed highly cost effective and has been described as the single most important intervention for HBV PMTCT, 91 underpinning international targets for worldwide coverage of 90% by 2030. 1 Modelling estimates suggest that >700,000 deaths in the 2020-2030 birth cohort could be prevented if this target were reached, with the greatest benefits in Africa. 16 For complete protection, HepB-BD vaccination needs to be timely (soon after birth, and definitely within 24 h), 16 and followed by two or three subsequent doses of vaccine, through the WHO Expanded Programme for Immunization (EPI). These subsequent doses are administered as a component of multivalent vaccines, typically starting at age 6 weeks, with a maximum recommended interval of 10 weeks between HepB-BD and the subsequent dose. 6,92,93 Local capacity for vaccine manufacture is improving and a treaty for the African Medicines Agency was ratified in November 2021, with the aim of enhancing the continent's regulatory and manufacturing independence. 94 HepB-BD vaccination: challenges and data gaps Logistical, financial and practical challenges There are substantial ongoing challenges in delivery, consistency and equity of supply, maintenance of the cold chain, timely administration (particularly to babies born outside healthcare settings, in indigenous communities and in rural areas), and political acceptance of the need for HepB-BD vaccination. 95 Despite global mandates, HepB-BD has not been consistently incorporated into regional or national guidance. In 2018, the coverage remained heterogeneous, globally estimated at 43%, and as low as 4-12% for the WHO Africa Region, where only 17/47 countries have formally adopted the strategy, compared to 75% in Latin America and >80% in the Western Pacific. 6,96-98 However, many African nations are making active progress towards introducing or enhancing HepB-BD vaccination, and there is a call for Gavi support to increase awareness of HBV and vaccination strategies, to train healthcare workers, and to enhance reach to infants born outside healthcare settings. 99 Although monovalent HBV vaccine is low cost, total costs to healthcare systems include infrastructure and personnel, and the cost to the patient can be substantially higher, for example costing 10USD in Cameroon, which is met by the individual. Coverage can be enhanced by strengthening primary care infrastructure and increasing the proportion of deliveries that occur with trained birth attendants and access to healthcare settings. More data are needed regarding the optimum timing of vaccine administration, but there is recognition that early administration is likely to be most beneficial (for example within 1 h of birth), 100 and that delay beyond 24 hours results in increasing risk.
There is increasing evidence for vaccine stability outside the cold chain for time-limited periods, 101 allowing wider reach without concerns about reduced efficacy, and with demonstrable health-economic benefits. 102 A 'controlled-temperature-chain' approach that allows vaccines to be held at ambient temperatures could replace the conventional stringent cold chain requirements. 103 Relicensing the monovalent HBV vaccine to provide this flexibility will be key to enhancing coverage. Most vaccines cannot be frozen, which may also be a logistical challenge.
HepB-BD delivery falls outside the conventionally agreed remit of the EPI, and the lack of centralised ownership of the BD programme is a potential barrier to implementation. Scale up of political commitment, focused investment, infrastructure, resources and education is now required, 104,105 which also depends on close integration with other aspects of maternal and infant care. 106 However, data from epidemiological studies and modelling indicate that even high coverage of infant vaccination (including HepB-BD) is not sufficient to reach the goal of a HBsAg prevalence of <0.1% in children aged 5 by 2030, 5,107 and this programme must be scaled up alongside other intervention strategies.

Vaccine resistance
The impact of vaccine escape mutations has not been robustly studied in many settings, but tracking of the emergence and clinical significance of these polymorphisms will be important. Current sequencing data suggest a low prevalence of individual vaccine escape mutations, but there is a lack of viral genetic data to represent many populations. 108 Unfortunately, HepB-BD vaccination is not infallible, as shown recently in Cameroon, where genetically engineered HBV A2 vaccines may not completely prevent MTCT from highly viraemic mothers infected with genotype E. 109 Stratification for HepB-BD vaccination A risk-stratified approach to delivery of HepB-BD vaccination has been considered, i.e. targeting the intervention to infants born to HBsAg+ mothers, rather than to all infants. This focused strategy theoretically has a higher chance of attaining high coverage of infants at-risk, and success is being demonstrated in some settings (mainly in Europe) in which there is low HBV endemicity, high rates of antenatal screening, and consistent implementation of maternal prophylaxis and timely HepB-BD vaccination; 110,111 use of this approach had been reported in 22 countries as of 2020. 112 However, there are risks that not all infants would be sufficiently protected by this targeted approach in settings with higher endemicity and/or challenges in deployment of screening and prophylaxis, and it does not cover infants at risk of transmission events from other household or family members. Given the challenges associated with maternal stratification and targeted administration of timely HepB-BD vaccination, the WHO has continued to endorse a universal approach. 6

HBIG: current evidence and interventions
HBIG is a post-exposure prophylaxis strategy for infants born to high-risk mothers, ideally based on high HBV DNA and/or HBeAg-positive status when this stratification is available. 113,114 A dose of 0.5 ml (at least 100 IU) is recommended for high-risk infants by many national and international guidelines ( Table 1), most of which refer (directly or indirectly) to WHO/CDC guidance first published in 1984. 115 HBIG requires storage at 2-8 C (but must not be frozen), 116 and has a limited shelf life (typically up to 1 year, but defined individually by suppliers). HBIG does not offer complete protection, particularly when the mother is HBeAg positive 57 and/or has high HBV DNA in the third trimester, 23,117-119 but in combination with timely HepB-BD it reduces the rate of perinatal transmission from >90% to <10% in high-risk pregnancies. 7,120 Follow-up HBIG doses may be advocated to further reduce the risk. 121 Neonatal post-exposure prophylaxis: challenges and data gaps Rationale for ongoing use of HBIG HBIG has been deemed cost effective (in addition to universal vaccination) for some populations, but there are often issues with accessibility and feasibility, particularly in low-income settings. 9,122,123 A mandate for HBIG administration can place a disproportionate burden on healthcare infrastructure and resources, and repeat HBIG doses are rarely practical due to the cost and difficulties in access. The theoretical possibility of transmission of prion disease has also contributed to scarcity of supply in some countries. 124 HBIG recommendations have not been revised since the advent of other PMTCT interventions, but there has been a lack of studies formally assessing the efficacy of HBIG-free regimens. 27 In the era of antiviral prophylaxis and HepB-BD vaccination, it is difficult to evaluate the impact of HBIG and signals for benefit frequently do not reach statistical significance. 125,126 However, data are now emerging. A study in Singapore found no benefit of HBIG for the infants of HBeAg-negative mothers, 127 while in Cambodia, HBIG offered no additional benefit if maternal TDF was started at least 4 weeks prior to delivery. 128 A meta-analysis of nine studies of HBeAg-negative mothers in Asia and Europe found no significant difference in rates of MTCT between neonates who received HepB-BD vaccination with or without HBIG. 129 In high-risk populations in Colombia, success in eliminating MTCT has been achieved without routine incorporation of HBIG. 95 A multicentre study is underway to determine the impact of omitting HBIG from prophylactic strategies. 130 To date, African populations have not been represented, and data are urgently needed as outcomes vary by setting and by viral genotype.
New guidelines must increasingly accommodate approaches that do not recommend routine HBIG use, focusing resources instead on strategies that are accessible, affordable, and sustainable. When HBIG is unavailable, maternal antiviral prophylaxis should be implemented as a priority, ideally supported by appropriate risk stratification, alongside timely HepB-BD vaccination.

Neonatal prophylaxis with antiviral therapy
In HIV-exposed infants, post-exposure prophylaxis is given with ART for 4-6 weeks, 131 setting a precedent for an approach that is widely deployed and accepted. To date, this strategy has not been explored for HBV, based on the alternative strategies of HepB-BD ± HBIG. However, recognising that timely delivery of vaccine and HBIG can be unreliable raises the question as to whether a defined course of antiviral prophylaxis could be a 'safety net' for selected high-risk HBV-exposed neonates. 109 Where failure to administer HepB-BD vaccination is due to birth outside healthcare facilities, chances of successful administration of antiviral prophylaxis may be similarly low.    HBIG, hepatitis B immunoglobulin; HepB-BD, hepatitis B birth dose vaccine; HBV, hepatitis B virus; HIV, human immunodeficiency virus; HPV, human papillomavirus; NA, nucleos(t)ide analogue; OPV, oral polio vaccine; PMTCT, prevention of mother-to-child transmission; POC, point of care; TAF, tenofovir alafenamide; TDF, tenofovir disoproxil fumarate; VL, viral load; XTC, 3TC (lamivudine) or FTC (emtricitabine). a In all categories, there is a need for better data that represent the populations at highest risk, especially recognising evidence gaps for the WHO Africa region. ** Interventions marked with a double-asterisk and in bold are of the highest priority. * Interventions marked with a single-asterisk are medium-ranked priorities. Furthermore, the current lack of availability of paediatric formulations and access to treatment for infants is a potential barrier. Research is therefore needed to explore the risk/benefits, alongside feasibility, acceptability and health-economic implications of this approach in different settings. The REVERT-B study is currently exploring the role for neonatal 3TC prophylaxis and early maternal tenofovir. 132

Delivery of enhanced PMTCT interventions
Enhanced approaches for HBV testing, antiviral prophylaxis and BD vaccination An enhanced PMTCT strategy should ideally (i) improve screening through enhanced access to both HBsAg and reflex HBV DNA testing including POCT and quantification; (ii) support wider maternal access to antiviral agents, with evidence to determine the role of longer courses (pre-and post-partum), broadened eligibility criteria, dual therapy, and use of TAF; (iii) deliver universal timely HBV-BD vaccination (or, in lowendemicity settings, provide targeted BD vaccination if robust screening and linkage to care is deployed), and (iv) consider a role for infant post-exposure prophylaxis with NA agents as an alternative to HBIG (Fig. 1, Table 3). Specific interventions may need to be tailored according to different settings, based on local epidemiology of infection, resources and infrastructure. However, to make meaningful and sustainable advances, simplified approaches and pursuit of a combination of strategies is likely to be important. In settings where resources are limited, increasing awareness, political support based on the scientific evidence, training and investment are crucial. Continuing to assume that costs can be met by national or regional healthcare systems, or by communities, families or individuals, will lead to a failure of deployment and worsening of already harsh health inequities. Accepting the status quo is inadequate: there is an urgent need for strategies that deliver interventions that are practical, accessible, affordable and effective.

Partnerships, integration and decentralisation
Strong partnerships with industry are required to secure consistent access to diagnostics, treatments and vaccines, and affordable HBeAg and HBV DNA tests are undoubtedly a central requirement for providing safe, equitable, and evidence-based PMTCT interventions. There is a pressing need for programmatic approaches with HBV PMTCT interventions that are decentralised and integrated into other health interventions, 10 including triple elimination, delivery of timely HepB-BD alongside other neonatal vaccines (neonatal BCG and oral polio), and provision of ongoing therapy for mothers and/or infants through established HIV infrastructure. 11,13,95,133,134 Additional Gavi investment to support universal HepB-BD access (in addition to existing provision of multivalent vaccinations through the EPI schedule), and Global Fund ownership of screening pregnant women under the triple elimination initiative, would strengthen systems. Timely BD vaccination depends on training of healthcare workers who are present at the time of birth, requiring taskshifting to involve different personnel from those who deliver the routine paediatric immunisation schedule, and outreach into communities which specifically considers risks associated with home-births, rural communities, and season of birth. 135,136 Financial incentives or subsidies could be provided to healthcare systems to boost coverage, while relaxation of strict cold chain requirements would support monovalent vaccine distribution in settings in which many births occur outside a healthcare setting.
Equity and representation PMTCT guidelines should account for differences between settings, and 'conditional' recommendations (as set out by the WHO) can be used to incorporate flexibility, such that approaches best serve the needs of specific populations. Enhanced data collection is required to determine baseline epidemiology and to track progress towards elimination targets (e.g. as described in Colombia). 95,137 Current evidence typically overlooks high-risk populations, for example rural communities, migrant populations, and marginalised groups for whom there are barriers to healthcare access; there is an urgent need to extend the reach of clinical, public health and research programmes to close these equity gaps. 81,138 The clinical and academic community, and patients and the public, should be uncompromising on demands for resources and intervention in order to drive high quality services with a clear focus on advocacy, universal access, and equity.
Education and community partnerships Stigma and discrimination associated with HBV infection are barriers to engagement, and PMTCT programmes for HBV have not been supported by appropriate education, awareness and community involvement. 139,140 In some populations, there is no local word for HBV infection, 141 highlighting a significant gap in awareness and knowledge. Practical implementation of PMTCT interventions is dependent on education of healthcare workers, and training birth attendants and local healthcare teams (e.g. community health extension workers). 136,142,143 Informing the population and healthcare workers creates demand for interventions, and provides crucial community engagement and advocacy. 144 There is typically a high level of acceptance of PMTCT interventions by women during pregnancy and postpartum, 145,146 and this population can be strong advocates for themselves and their infants. PMTCT can be considered a family intervention, as identifying HBsAg-positive status in pregnancy also enables linkage to HBV care post-partum for the mother while providing the opportunity to screen (±vaccinate) partners, and other family members. These interventions function to break ongoing cycles of infection in families and communities.

Conclusion
HBV PMTCT is a cornerstone of strategies to reduce the incidence and prevalence of HBV infection in children, lowering the population reservoir of virus (with an impact on reducing transmission), and reducing longer-term burden of disease, including cancer incidence. Tools are in place to abrogate PMTCT, but these have not been deployed adequately or consistently to date. There is now an urgent need to scale up existing interventions while tackling evidence gaps to optimise practice, reduce inequity and meet global targets for HBV elimination.

Conflict of interest
CWS has received speaker fees from GILEAD Sciences and Abbott. CP received research funding and is a speaker for Gilead. PCM supervises a doctoral student with funding support from GSK. SH has received funding from Gilead for HCV Micro-elimination programs. HR has received research funding and speaker's honoraria from Gilead Sciences, AbbVie and Pfizer and is a board member of the CDA foundation. SW has received research funding from Gilead Sciences, and honoraria from Prime Inc, and is on the Board of Directors for the Hepatitis B Foundation, World Hepatitis Alliance, and serves in the patient advisory group and HBV special interest group for the AASLD. FR has received consulting fees from Sanofi Pasteur. GD has received consulting fees and speaker fees from Gilead Sciences, has participated on Data Safety Monitoring Board/Advisory Board for janssen, Glaxo Smith Kline, Arbutus, Aligos, Vir and has roles in the National Medical Research Council Singapore and the World Health Organisation Pediatric Working Group on Viral Hepatitis.
Please refer to the accompanying ICMJE disclosure forms for further details.